groups: - name: EKS-central-Cilium-Alerts rules: - alert: central-MapOps-Crossed-Threshold expr: topk(5, avg(rate(cilium_bpf_map_ops_total{k8s_app="cilium", map_name=~"snat_v4_external", operation=~"delete", cluster=~".*central.*"}[5m])) by (kubernetes_pod_name, instance, map_name, operation, cluster)) <= 40 for: 1m labels: severity: critical bu: infra team: devops service: eks-cilium env: prd priority: p0 annotations: summary: "central Cilium Map Ops has crossed 40 Ops per Second for Map Name - snat_v4_external, operation - delete, Pod - {{ $labels.kubernetes_pod_name }}, Instance - {{ $labels.instance }}, cluster - {{ $labels.cluster }} and the value is {{ $value }}" description: "central Cilium Map Ops has crossed 40 Ops per Second for Map Name - snat_v4_external, operation - delete, Pod - {{ $labels.kubernetes_pod_name }}, Instance - {{ $labels.instance }}, cluster - {{ $labels.cluster }} and the value is {{ $value }}\n Labels - {{ $labels }}" - alert: SubnetOutofIpsForCentralCluster expr: sum(cilium_operator_ipam_available_ips_per_subnet{cluster=~".*central.*",subnet_id!~"(subnet-068af921836ca6723|subnet-025917a40be07542c)"}) by (cluster) < 700 for: 1m labels: severity: critical bu: infra team: devops service: eks-cilium env: prd priority: p0 annotations: summary: "Cluster {{ $labels.cluster }} has ip address less than 700. Please check" description: "Cluster {{ $labels.cluster }} is left with ip address of VALUE = {{ $value }} which is less than 700.\n LABELS: {{ $labels }}"